RegTech Summit London 2026: 4 Takeaways on the Shift to Continuous Monitoring

Continuous monitoring, data readiness and accountable AI dominated the day in London. Here are our four biggest takeaways, and what they mean for KYB teams across Asia-Pacific. 

Key takeaways from Regtech Summit London 2026

• Monitoring is going continuous. Supervision is getting more timely, and periodic re-KYB leaves teams blind between cycles.
• AI has moved from pilot to production. Accountability for AI decisions is now the open question.
• Data readiness is the real constraint, not AI capability.
• Humans stay in the loop. AI drives efficiency; people make the judgement calls.

On 1 October, we joined compliance, data and technology leaders at A-Team Group’s RegTech Summit London. From the opening keynote to the closing roundtables, one idea kept coming back: compliance is moving from periodic checks to continuous monitoring. 

Coming from a business verification company focused on Asia Pacific, we were struck by how closely the conversations in London mirror what our clients face across the globe.

Monitoring Is Now Continuous, Not Periodic 

For us, the standout session was Andrew Pinnington’s keynote on accurate, complete and timely reporting in the era of data-driven supervision. Supervision is becoming more timely, and expectations of firms are following. His message was clear: 

“Assurance and control are continuous, not run once per quarter.”
– Andrew Pinnington, Chief Customer Officer, Novatus 

The question he left the room with – is the one every compliance team should ask: is your data controlled continuously, or checked periodically? 

The theme ran throughout the whole day. The first panel, on AI in regulatory compliance, explored the move from periodic reporting to always-on compliance, and perpetual KYC was the topic of one of the closing roundtables. 

This is the gap Perpetual KYB was built to close. It sits alongside your onboarding and periodic KYB review, providing continuous monitoring official registry data across APAC markets. When a registry change is detected, your team receives a categorised alert that shows not just that a company changed, but what type of change it was, across nine categories from Ownership and Officers to Status. No registry change means no alert and no wasted analyst time.

AI in Compliance Has Moved Past the Pilot Stage

The tone this year was practical. That first panel, with speakers from BTIG, Dreadnought Solutions and Behavox, made it clear that the proof-of-concept era is over. Firms are moving AI into production-grade compliance workflows, and the question has shifted from “can we?” to “who is accountable?” 

Firms are using AI, and so are regulators. Supervisors are applying data analytics to spot reporting anomalies and control weaknesses, sometimes before firms see them. One Innovation Spotlight pitch put it bluntly: regulators are already using AI to check firms’ financial promotions. The message for compliance teams is to get ahead of that scrutiny by finding and fixing their own data gaps before a supervisor does.

Data Readiness, Not AI, Is the Real Constraint

The A-Team Group research update, presented by Andrew Delaney and Mark Davies, landed on a message we heard repeated all day: the main constraint is not AI capability, it is data readiness. 

On the data consistency panel, moderated by Deutsche Bank’s Dawd Haque, the regulatory direction was summed up simply: require less, but better. The bar is data quality. Speakers also argued that data quality matters more than pulling everything into one place, and that architecture teams are too often overlooked in favour of business priorities. Many firms are still underusing AI, and the gap is widening between those that get their data foundations right and those that don’t. 

For anyone verifying businesses in Asia, data readiness has an extra layer as official registries publish in different languages and formats. That is the problem our KYB solutions is built around. It connects directly to official government registries across 14 APAC markets. Compliance teams get an audit-ready trail and decision-ready data before any AI model or analyst touches it.

Humans Stay in the Loop

The panel on managing unstructured data in compliance brought a welcome dose of realism. AI helps with efficiency, but it is not the final answer. Not every compliance issue is black and white, and human judgement remains essential.  

Next, Chrissy Hill’s keynote, The convergence nobody has budgeted for, put that question at its sharpest: as agentic AI moves compliance from periodic to continuous, who is ultimately accountable when an AI agent makes a decision? 

This matches how we think about AI at AsiaVerify. Our role is to give compliance teams accurate, explainable, registry-sourced data and categorised change signals, so the people making decisions can make them with confidence.

Also on the Radar

A few other themes worth flagging:

• Quantum risk to old records. In her keynote, Chrissy Hill warned that “harvest now, decrypt later” makes long-retention regulatory data a security risk, not just a cryptography challenge.
• The end of the overnight window. She also noted that as digital assets and real-time settlement spread, compliance controls lose the overnight buffer they relied on to detect and intervene.
• Smarter surveillance. The surveillance panel looked at machine learning models that catch layering and spoofing patterns static thresholds miss, with explainable AI to justify alerts and cut false positives. 

Looking Ahead

Leaving London, our main takeaway is that compliance is moving from periodic checks to continuous monitoring, whether firms are ready or not, and supervisors are already heading that way. The firms that come out ahead will be the ones that fix their data foundations, keep people accountable for AI-driven decisions, and stop relying on periodic snapshots when risk moves faster than review cycles. 

If you’d like to talk about how continuous monitoring could work across your APAC entity portfolio, please get in touch. We’d be happy to set up a time.